Return a 404 if no valid thumbnail is found. (#9163)

If no thumbnail of the requested type exists, return a 404 instead
of erroring. This doesn't quite match the spec (which does not define
what happens if no thumbnail can be found), but is consistent with
what Synapse already does.

4 files changed, 183 insertions, 82 deletions

diff --git a/changelog.d/9163.bugfix b/changelog.d/9163.bugfix
new file mode 100644
index 0000000000..c51cf6ca80
--- /dev/null
+++ b/changelog.d/9163.bugfix
@@ -0,0 +1 @@
+Fix a long-standing bug where Synapse would return a 500 error when a thumbnail did not exist (and auto-generation of thumbnails was not enabled).
diff --git a/synapse/rest/media/v1/_base.py b/synapse/rest/media/v1/_base.py
index 31a41e4a27..f71a03a12d 100644
--- a/synapse/rest/media/v1/_base.py
+++ b/synapse/rest/media/v1/_base.py
@@ -300,6 +300,7 @@ class FileInfo:
         thumbnail_height (int)
         thumbnail_method (str)
         thumbnail_type (str): Content type of thumbnail, e.g. image/png
+        thumbnail_length (int): The size of the media file, in bytes.
     """
 
     def __init__(
@@ -312,6 +313,7 @@ class FileInfo:
         thumbnail_height=None,
         thumbnail_method=None,
         thumbnail_type=None,
+        thumbnail_length=None,
     ):
         self.server_name = server_name
         self.file_id = file_id
@@ -321,6 +323,7 @@ class FileInfo:
         self.thumbnail_height = thumbnail_height
         self.thumbnail_method = thumbnail_method
         self.thumbnail_type = thumbnail_type
+        self.thumbnail_length = thumbnail_length
 
 
 def get_filename_from_headers(headers: Dict[bytes, List[bytes]]) -> Optional[str]:
diff --git a/synapse/rest/media/v1/thumbnail_resource.py b/synapse/rest/media/v1/thumbnail_resource.py
index d6880f2e6e..d653a58be9 100644
--- a/synapse/rest/media/v1/thumbnail_resource.py
+++ b/synapse/rest/media/v1/thumbnail_resource.py
@@ -16,7 +16,7 @@
 
 
 import logging
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Any, Dict, List, Optional
 
 from twisted.web.http import Request
 
@@ -106,31 +106,17 @@ class ThumbnailResource(DirectServeJsonResource):
             return
 
         thumbnail_infos = await self.store.get_local_media_thumbnails(media_id)
-
-        if thumbnail_infos:
-            thumbnail_info = self._select_thumbnail(
-                width, height, method, m_type, thumbnail_infos
-            )
-
-            file_info = FileInfo(
-                server_name=None,
-                file_id=media_id,
-                url_cache=media_info["url_cache"],
-                thumbnail=True,
-                thumbnail_width=thumbnail_info["thumbnail_width"],
-                thumbnail_height=thumbnail_info["thumbnail_height"],
-                thumbnail_type=thumbnail_info["thumbnail_type"],
-                thumbnail_method=thumbnail_info["thumbnail_method"],
-            )
-
-            t_type = file_info.thumbnail_type
-            t_length = thumbnail_info["thumbnail_length"]
-
-            responder = await self.media_storage.fetch_media(file_info)
-            await respond_with_responder(request, responder, t_type, t_length)
-        else:
-            logger.info("Couldn't find any generated thumbnails")
-            respond_404(request)
+        await self._select_and_respond_with_thumbnail(
+            request,
+            width,
+            height,
+            method,
+            m_type,
+            thumbnail_infos,
+            media_id,
+            url_cache=media_info["url_cache"],
+            server_name=None,
+        )
 
     async def _select_or_generate_local_thumbnail(
         self,
@@ -276,26 +262,64 @@ class ThumbnailResource(DirectServeJsonResource):
         thumbnail_infos = await self.store.get_remote_media_thumbnails(
             server_name, media_id
         )
+        await self._select_and_respond_with_thumbnail(
+            request,
+            width,
+            height,
+            method,
+            m_type,
+            thumbnail_infos,
+            media_info["filesystem_id"],
+            url_cache=None,
+            server_name=server_name,
+        )
 
+    async def _select_and_respond_with_thumbnail(
+        self,
+        request: Request,
+        desired_width: int,
+        desired_height: int,
+        desired_method: str,
+        desired_type: str,
+        thumbnail_infos: List[Dict[str, Any]],
+        file_id: str,
+        url_cache: Optional[str] = None,
+        server_name: Optional[str] = None,
+    ) -> None:
+        """
+        Respond to a request with an appropriate thumbnail from the previously generated thumbnails.
+
+        Args:
+            request: The incoming request.
+            desired_width: The desired width, the returned thumbnail may be larger than this.
+            desired_height: The desired height, the returned thumbnail may be larger than this.
+            desired_method: The desired method used to generate the thumbnail.
+            desired_type: The desired content-type of the thumbnail.
+            thumbnail_infos: A list of dictionaries of candidate thumbnails.
+            file_id: The ID of the media that a thumbnail is being requested for.
+            url_cache: The URL cache value.
+            server_name: The server name, if this is a remote thumbnail.
+        """
         if thumbnail_infos:
-            thumbnail_info = self._select_thumbnail(
-                width, height, method, m_type, thumbnail_infos
+            file_info = self._select_thumbnail(
+                desired_width,
+                desired_height,
+                desired_method,
+                desired_type,
+                thumbnail_infos,
+                file_id,
+                url_cache,
+                server_name,
             )
-            file_info = FileInfo(
-                server_name=server_name,
-                file_id=media_info["filesystem_id"],
-                thumbnail=True,
-                thumbnail_width=thumbnail_info["thumbnail_width"],
-                thumbnail_height=thumbnail_info["thumbnail_height"],
-                thumbnail_type=thumbnail_info["thumbnail_type"],
-                thumbnail_method=thumbnail_info["thumbnail_method"],
-            )
-
-            t_type = file_info.thumbnail_type
-            t_length = thumbnail_info["thumbnail_length"]
+            if not file_info:
+                logger.info("Couldn't find a thumbnail matching the desired inputs")
+                respond_404(request)
+                return
 
             responder = await self.media_storage.fetch_media(file_info)
-            await respond_with_responder(request, responder, t_type, t_length)
+            await respond_with_responder(
+                request, responder, file_info.thumbnail_type, file_info.thumbnail_length
+            )
         else:
             logger.info("Failed to find any generated thumbnails")
             respond_404(request)
@@ -306,67 +330,117 @@ class ThumbnailResource(DirectServeJsonResource):
         desired_height: int,
         desired_method: str,
         desired_type: str,
-        thumbnail_infos,
-    ) -> dict:
+        thumbnail_infos: List[Dict[str, Any]],
+        file_id: str,
+        url_cache: Optional[str],
+        server_name: Optional[str],
+    ) -> Optional[FileInfo]:
+        """
+        Choose an appropriate thumbnail from the previously generated thumbnails.
+
+        Args:
+            desired_width: The desired width, the returned thumbnail may be larger than this.
+            desired_height: The desired height, the returned thumbnail may be larger than this.
+            desired_method: The desired method used to generate the thumbnail.
+            desired_type: The desired content-type of the thumbnail.
+            thumbnail_infos: A list of dictionaries of candidate thumbnails.
+            file_id: The ID of the media that a thumbnail is being requested for.
+            url_cache: The URL cache value.
+            server_name: The server name, if this is a remote thumbnail.
+
+        Returns:
+             The thumbnail which best matches the desired parameters.
+        """
+        desired_method = desired_method.lower()
+
+        # The chosen thumbnail.
+        thumbnail_info = None
+
         d_w = desired_width
         d_h = desired_height
 
-        if desired_method.lower() == "crop":
+        if desired_method == "crop":
+            # Thumbnails that match equal or larger sizes of desired width/height.
             crop_info_list = []
+            # Other thumbnails.
             crop_info_list2 = []
             for info in thumbnail_infos:
+                # Skip thumbnails generated with different methods.
+                if info["thumbnail_method"] != "crop":
+                    continue
+
                 t_w = info["thumbnail_width"]
                 t_h = info["thumbnail_height"]
-                t_method = info["thumbnail_method"]
-                if t_method == "crop":
-                    aspect_quality = abs(d_w * t_h - d_h * t_w)
-                    min_quality = 0 if d_w <= t_w and d_h <= t_h else 1
-                    size_quality = abs((d_w - t_w) * (d_h - t_h))
-                    type_quality = desired_type != info["thumbnail_type"]
-                    length_quality = info["thumbnail_length"]
-                    if t_w >= d_w or t_h >= d_h:
-                        crop_info_list.append(
-                            (
-                                aspect_quality,
-                                min_quality,
-                                size_quality,
-                                type_quality,
-                                length_quality,
-                                info,
-                            )
+                aspect_quality = abs(d_w * t_h - d_h * t_w)
+                min_quality = 0 if d_w <= t_w and d_h <= t_h else 1
+                size_quality = abs((d_w - t_w) * (d_h - t_h))
+                type_quality = desired_type != info["thumbnail_type"]
+                length_quality = info["thumbnail_length"]
+                if t_w >= d_w or t_h >= d_h:
+                    crop_info_list.append(
+                        (
+                            aspect_quality,
+                            min_quality,
+                            size_quality,
+                            type_quality,
+                            length_quality,
+                            info,
                         )
-                    else:
-                        crop_info_list2.append(
-                            (
-                                aspect_quality,
-                                min_quality,
-                                size_quality,
-                                type_quality,
-                                length_quality,
-                                info,
-                            )
+                    )
+                else:
+                    crop_info_list2.append(
+                        (
+                            aspect_quality,
+                            min_quality,
+                            size_quality,
+                            type_quality,
+                            length_quality,
+                            info,
                         )
+                    )
             if crop_info_list:
-                return min(crop_info_list)[-1]
-            else:
-                return min(crop_info_list2)[-1]
-        else:
+                thumbnail_info = min(crop_info_list)[-1]
+            elif crop_info_list2:
+                thumbnail_info = min(crop_info_list2)[-1]
+        elif desired_method == "scale":
+            # Thumbnails that match equal or larger sizes of desired width/height.
             info_list = []
+            # Other thumbnails.
             info_list2 = []
+
             for info in thumbnail_infos:
+                # Skip thumbnails generated with different methods.
+                if info["thumbnail_method"] != "scale":
+                    continue
+
                 t_w = info["thumbnail_width"]
                 t_h = info["thumbnail_height"]
-                t_method = info["thumbnail_method"]
                 size_quality = abs((d_w - t_w) * (d_h - t_h))
                 type_quality = desired_type != info["thumbnail_type"]
                 length_quality = info["thumbnail_length"]
-                if t_method == "scale" and (t_w >= d_w or t_h >= d_h):
+                if t_w >= d_w or t_h >= d_h:
                     info_list.append((size_quality, type_quality, length_quality, info))
-                elif t_method == "scale":
+                else:
                     info_list2.append(
                         (size_quality, type_quality, length_quality, info)
                     )
             if info_list:
-                return min(info_list)[-1]
-            else:
-                return min(info_list2)[-1]
+                thumbnail_info = min(info_list)[-1]
+            elif info_list2:
+                thumbnail_info = min(info_list2)[-1]
+
+        if thumbnail_info:
+            return FileInfo(
+                file_id=file_id,
+                url_cache=url_cache,
+                server_name=server_name,
+                thumbnail=True,
+                thumbnail_width=thumbnail_info["thumbnail_width"],
+                thumbnail_height=thumbnail_info["thumbnail_height"],
+                thumbnail_type=thumbnail_info["thumbnail_type"],
+                thumbnail_method=thumbnail_info["thumbnail_method"],
+                thumbnail_length=thumbnail_info["thumbnail_length"],
+            )
+
+        # No matching thumbnail was found.
+        return None
diff --git a/tests/rest/media/v1/test_media_storage.py b/tests/rest/media/v1/test_media_storage.py
index ae2b32b131..a6c6985173 100644
--- a/tests/rest/media/v1/test_media_storage.py
+++ b/tests/rest/media/v1/test_media_storage.py
@@ -202,7 +202,6 @@ class MediaRepoTests(unittest.HomeserverTestCase):
 
         config = self.default_config()
         config["media_store_path"] = self.media_store_path
-        config["thumbnail_requirements"] = {}
         config["max_image_pixels"] = 2000000
 
         provider_config = {
@@ -313,15 +312,39 @@ class MediaRepoTests(unittest.HomeserverTestCase):
         self.assertEqual(headers.getRawHeaders(b"Content-Disposition"), None)
 
     def test_thumbnail_crop(self):
+        """Test that a cropped remote thumbnail is available."""
         self._test_thumbnail(
             "crop", self.test_image.expected_cropped, self.test_image.expected_found
         )
 
     def test_thumbnail_scale(self):
+        """Test that a scaled remote thumbnail is available."""
         self._test_thumbnail(
             "scale", self.test_image.expected_scaled, self.test_image.expected_found
         )
 
+    def test_invalid_type(self):
+        """An invalid thumbnail type is never available."""
+        self._test_thumbnail("invalid", None, False)
+
+    @unittest.override_config(
+        {"thumbnail_sizes": [{"width": 32, "height": 32, "method": "scale"}]}
+    )
+    def test_no_thumbnail_crop(self):
+        """
+        Override the config to generate only scaled thumbnails, but request a cropped one.
+        """
+        self._test_thumbnail("crop", None, False)
+
+    @unittest.override_config(
+        {"thumbnail_sizes": [{"width": 32, "height": 32, "method": "crop"}]}
+    )
+    def test_no_thumbnail_scale(self):
+        """
+        Override the config to generate only cropped thumbnails, but request a scaled one.
+        """
+        self._test_thumbnail("scale", None, False)
+
     def _test_thumbnail(self, method, expected_body, expected_found):
         params = "?width=32&height=32&method=" + method
         channel = make_request(